By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Microsoft Discussion - Microsoft: Xbox Series S and X only next-gen consoles with full RDNA 2 feature set

Oh boy...



Around the Network
Hynad said:
shikamaru317 said:

https://www.pczone.co.uk/hardware-vs-software-ray-tracing-whats-the-difference/

Both the PS5 and Series X rely on their CUs to achieve Ray Tracing. RDNA2 doesn’t have dedicated RT and Tensor cores like Nvidia’s RTX approach.

Not entirely accurate.
AMD has included a dedicated "Ray Accelerator" in every single CU with RDNA2. - It's essentially a fixed function Ray Tracing "core".
AMD may be using FP16 compute (Leveraging Rapid Packed Math) instead of tensor cores considering how much better their GPU designs generally are with general purpose asynchronous compute anyway.

We need to remember what a CU is... It's essentially a grouping of processing cores/functional blocks, some specialized, some general purpose.

chakkra said:
Pemalite said:

It's not just about performance, that is a fallacious line of thinking.

Often a feature gets introduced not because of performance reasons, but because it may also reduce development burden on game developers.
Remember what runs on top of graphics hardware... Games.
nVidia and AMD both work closely with developers and various API's to extract as much ease-of-development as possible with various development tools and optimizations.

I.E. Case in point TressFX. - It allows for advanced simulation of hair, grass and fur by leveraging various middle-ware libraries that just happened to be GPU accelerated. - It greatly expedited development time, but enabling it? There was certainly an impact to performance and efficiency, but that was worth the trade-off for the visual gains.

So no. It's not just about performance, it's often just about reducing development time and increasing visuals.

Okay, let's see:

Sampler Feedback Streaming (SFS) – Is a feature that allows games to load into memory, with fine granularity, only the portions of textures that the GPU needs for a scene, as it needs it. This enables far better memory utilization for textures, which is important given that every 4K texture consumes 8MB of memory. = Performance

Variable Rate Shading (VRS) – Increases GPU efficiency by concentrating shader work where it’s most needed and reducing shader work in areas where it won’t be noticeable. With minimal developer effort, VRS significantly improves GPU performance resulting in more stable and higher resolutions and frame rates with no perceptible loss in visual quality. = Performance

Mesh Shading – Mesh shading will enable developers to dramatically improve the performance and image quality when rendering a substantial number of complex objects in a scene. As an example, mesh shaders could enable the player to experience asteroid belts and fields of flowers in more pristine detail without seeing a loss in performance. = Performance

DirectStorage – Is an all new I/O system designed specifically for gaming to unleash the full performance of the SSD and hardware decompression. Modern games perform asset streaming in the background to continuously load the next parts of the world while you play, and DirectStorage can reduce the CPU overhead for these I/O operations from multiple cores to taking just a small fraction of a single core; thereby freeing considerable CPU power for the game to spend on areas like better physics or more NPCs in a scene. = Performance

https://news.xbox.com/en-us/2020/03/16/xbox-series-x-glossary/

Now let's see what Anandtech have to say about these features in specific:

"Meanwhile variable rate shading and mesh shaders are going to be less visible to end users, but they offer tangible performance improvements, and in the case of mesh shaders will eventually dramatically alter the geometry pipeline for games designed with mesh shaders as a baseline feature. Finally, sampler feedback will allow game developers to get a better idea of what textures and texel blocks within those textures are being used, allowing developers to better manage what assets are in VRAM and what needs to be pre-loaded."

https://www.anandtech.com/show/16202/amd-reveals-the-radeon-rx-6000-series-rdna2-starts-at-the-highend-coming-november-18th/2

None of that refutes anything I have said prior.



--::{PC Gaming Master Race}::--

Pemalite said:
Hynad said:

Both the PS5 and Series X rely on their CUs to achieve Ray Tracing. RDNA2 doesn’t have dedicated RT and Tensor cores like Nvidia’s RTX approach.

Not entirely accurate.
AMD has included a dedicated "Ray Accelerator" in every single CU with RDNA2. - It's essentially a fixed function Ray Tracing "core".
AMD may be using FP16 compute (Leveraging Rapid Packed Math) instead of tensor cores considering how much better their GPU designs generally are with general purpose asynchronous compute anyway.

We need to remember what a CU is... It's essentially a grouping of processing cores/functional blocks, some specialized, some general purpose.

I wasn’t aware of that but I am not surprised they did this, since it makes sense.

The point I was making still stands though. Both the PS5 and Series X use their CUs to provide hardware accelerated ray tracing.



Pemalite said:
chakkra said:

If we forget for a moment about PS5 vs XSX (which is the part that makes everybody shut their doors) and just focus on RDNA2 vs RDNA1, there has been a few articles outlining the differences between them.

"Ever since AMD announced the RDNA2 architecture, they have reiterated a singular goal: they wanted to achieve a 50% jump in perf-per-watt over RDNA1. And that they would accomplish it entirely with architectural improvements, not process improvements..."

"Along with numerous optimizations to the power efficiency of their GPU architecture, RDNA2 also includes a much-needed update to the graphics side of AMD’s GPU architecture. RDNA (1), though a massive replumbing of the core compute architecture, did not include any graphics feature upgrades. As a result, AMD only offered a DirectX feature level 12_1 feature set – the same as the Radeon RX Vega series – at a time when NVIDIA was offering ray tracing and the other features that have since become DirectX 12 Ultimate"

https://www.anandtech.com/show/16202/amd-reveals-the-radeon-rx-6000-series-rdna2-starts-at-the-highend-coming-november-18th/2

 I am talking low-level stuff not marketing fluff laid on top, that is for those who don't delve into the hardware nitty gritty.

chakkra said:

And about Primitive Shaders and Mesh Shaders being exactly the same, not quite. 

"A Mesh shader is a new type of shader that combines vertex and primitive processing. VS, HS, DS, and GS shader stages are replaced with Amplification Shader and Mesh Shader. Roughly, Mesh shaders replace VS+GS or DS+GS shaders and Amplification shaders replace VS+HS."

https://microsoft.github.io/DirectX-Specs/d3d/MeshShader.html

"Mesh shaders represent a radical simplification of the geometry pipeline. With a mesh shader enabled, all the shader stages and fixed-function features described above are swept away. Instead, we get a clean, straightforward pipeline using a compute-shader-like programming model. Importantly, this new pipeline is both highly flexible—enough to handle the existing geometry tasks in a typical game, plus enable new techniques that are challenging to do on the GPU today— it looks like it should be quite performance-friendly, with no apparent architectural barriers to efficient GPU execution."

https://www.starcitizen.gr/2642867-2/

They are the same from a user standpoint. One is essentially a marketing term from nVidia, another is a marketing term from AMD.

With AMD's set-up you have to assemble the input of a pre-defined format like vertices+vertex indices sequentially, where-as with a mesh-shader it's entirely defined by the user and is thus not bounded by the assembly stage.

From a developer/hardware point of view AMD's approach with in-driver shader transformation, the advantages are limited compared to full mesh shader support, as programmability is sacrificed.

But for all intents and purposes, they are the same.
Are there more differences? Of course. But they set out to offer the same result, ends users aren't going to care about the finer points.

AsGryffynn said:

Not necessarily. Most of the loss in power does translate into... well... heat. If a lighter supply is feeding a more powerful machine, there's less heat generation from the PSU. In other words... if the PS5 does have a more powerful unit, it might be generating more heat. 

I am not saying they are 400w units, just saying that the wattage a unit might have stickered onto the side is irrelevant.

AsGryffynn said:

Not necessarily. Most of the loss in power does translate into... well... heat. If a lighter supply is feeding a more powerful machine, there's less heat generation from the PSU. In other words... if the PS5 does have a more powerful unit, it might be generating more heat. 

I already touched on thermodynamics and energy conversion efficiency. - Which includes heat.

And it's not just about a more powerful PSU vs weaker PSU. It's down to energy conversion efficiency, higher efficiency means less wasted heat during conversion.

drkohler said:

Actually, the exact opposite could be true.

If a 350W ps operates in its "comfort zone", and a 320W ps operates slightly above its "comfort zone", then the latter ps generates more heat.

However, the whole thing depends on the voltage regulator circuitry. Contrary to what you seem to think, the ps does not feed "the machine", it feeds the voltage regulator circuitry (where much more heat is generated than in the ps). The vrc is the place where money vs heat is traded at design time (more phases = less heat but higher costs).

Nothing has been revealed about where the ps work or how the vrcs are built so any discussion about "this comsole is better than that one" is pointless.

My point is from the very beginning is that the hardware needs to actually be tested before anything definitive can be asserted.

AsGryffynn said:

I'm going on a limb and saying they are both operating at normal capacity. 

Well, at that point we're dealing with internal components so our guess will pretty much be moot unless we get someone like Louis Rossmann to tear both consoles down and look at them on a component level. I do assume that they are going to be fairly similar if not identical however. If this is the case, then the ball falls back on the PSU's energy loss and not the heat generation of the SoC (there's a good reason why the CPU and PSU get the cooling on the spot, but then again, we're dealing with APUs, so presumably the heat is going somewhere else. 

As years have gone by and Power supplies have gotten more efficient... Many PC PSU's won't actually initiate their fan until the unit exceeds a certain temperature threshold anyway.

Traditionally PC's have also used the PSU to exhaust the cases air as well.

The thing is we have no clue whatsoever what the efficiency for these two is... do we? 



chakkra said:
hinch said:

This is marketing of course they would say they have the 'full RNDA 2' feature set they created half of them in collaboration with AMD. Sony simply can't call it half of those since they've trademarked some of the names with Direct X. In addition people focusing on RDNA 1, 2, 3 feature as technobabble to one up each other is wrong. Both consoles are RDNA 2 based consoles one is just a but more customised and the other is designed with the 'full' features that work on DirectX.

PS5 does indeed have 'mesh shading' and 'VRR'. Sony has removed the standard VRR and customised the Geometry Engine for their needs.


FYI this guy is a developer and has a PS5 dev kit. This is what people have been talking about when saying 'RDNA 3' feature in PS5. Future AMD GPU's will most likely have a variation of this technology.

Both have the same or similar features sets. Different names and different way of going about it. Also a bit misleading of Microsoft since Series X/S don't have Infinity Cache, but there you go. Enjoy your consoles, both will be powerful machines. stop sweating the details :p

That is certainly an interesting read. I just can't help but wonder why Sony decided to leave the technology inside their chip out of their marketing strategy, and especially out of Cerny's presentation. You would think that saying "Hey, our chip might be smaller but it has all these cool technologies" sounds better than saying "smaller chips are nimble."

I don't think most people care to be honest about all these terms. Most gamers don't know jargon  and majority won't pay attention. This kind of stuff won't generate sales. How many people who listened to cerny  speak got much out if it? Probably only a few. Most people just want to hear stuff about games or see what the thing looks like. 



Around the Network
KratosLives said:
chakkra said:

That is certainly an interesting read. I just can't help but wonder why Sony decided to leave the technology inside their chip out of their marketing strategy, and especially out of Cerny's presentation. You would think that saying "Hey, our chip might be smaller but it has all these cool technologies" sounds better than saying "smaller chips are nimble."

I don't think most people care to be honest about all these terms. Most gamers don't know jargon  and majority won't pay attention. This kind of stuff won't generate sales. How many people who listened to cerny  speak got much out if it? Probably only a few. Most people just want to hear stuff about games or see what the thing looks like. 

And yet he spent a sizeable amount of time talking about variable clock rates, SSD, and 3D audio. That was certainly an interesting choice of which "jargons" should be at the forefront.



shikamaru317 said:
LudicrousSpeed said:
I don’t think Ubi has confirmed Valhalla details for SX yet. It very well could also be upscaled or dynamic. But maybe they have and I just missed it.

From what I'm reading, Sony created some confusion with the wording on the blog post about how AC runs on PS5 and XSX. When asked by a DualShockers journalist to clarify, an Ubisoft representative said that AC Valhalla is "native 4k 60 fps on XSX" and "4K at 60 fps on PS5 (upscaled 4K)".

Oh ok, cool. Apparently Watch Dogs is slightly above 1440p on SX with RT. Makes me wonder what the PS5 and SS versions will run. Probably the same thing for PS5, 900p low point for SS.

Yasnoozea supposedly runs at 4k60 on PS5 but only 4k30 on SX. I think I’m waiting for DF to test all this before I accept it, not that it matters much anyway. Imho RT and frame rates are the most important factor for next gen.



Considering Demon's Souls is the best looking next gen title, the Xbox Series X/S having full RDNA2 won't matter, until the Series X has a title that looks heads and shoulders better than any PS5 titles.

Otherwise, its a minor advantage that doesn't compare to the potential of the SSD/ IO throughput, with near instant load times being the tip of the iceberg.



If Watch Dogs lives up to the effects video they put out today, it’s easily the most impressive thing I have seen for next gen especially when you consider the size and scope of the game. It should look great on all next gen platforms, SS maybe withstanding, who knows what the effects will look like there.



PotentHerbs said:
Considering Demon's Souls is the best looking next gen title, the Xbox Series X/S having full RDNA2 won't matter, until the Series X has a title that looks heads and shoulders better than any PS5 titles.

Otherwise, its a minor advantage that doesn't compare to the potential of the SSD/ IO throughput, with near instant load times being the tip of the iceberg.

demon souls

scorn

No.